FIELD OF THE INVENTIONThe present invention relates generally to spectroscopy methods and apparatus, and more particularly to microwave time delay spectroscopy methods and apparatus for remote interrogation of biological targets.
DESCRIPTION OF THE PRIOR ART AND PRIOR ART STATEMENTThe essential requirements for a microwave measurement technique which is to be used to interrogate remote targets are that it be non-ambiguous for biologically relevant electrical path lengths, and that it be capable of discriminating between biologically relevant path length variations. Further, in order to have practical utility, the measurement technique should be easy to implement, and the data provided thereby must be readily interpretable.
When electromagnetic energy in the microwave region propagates through a biological target, the phase velocity and absorption thereof are a function of the permittivity, or dielectric constant, of the medium. Since the dielectric constant of a biological medium is a function of the composition and functional state of the tissue, the measurement of the total time delay and attenuation in microwave energy propagated through a biological target can theoretically be used to characterize the type, functional state, and thickness of the tissue through which the wave length travels. This capability is to be contrasted with the capability of X-ray interrogation techniques, where it is not possible to characterize tissue types by propagation time measurements. However, any time delay measurement technique which is utilized must provide unambiguous information between multiple paths whose differential propagation time is quite small.
For a number of reasons, localization and characterization of the dielectric properties of nonhomogeneous biological targets cannot be easily accomplished using only single frequency continuous wave (CW) measurements of reflection and transmission coefficients. One important reason is that the energy reaching the receiving attenna cannot be assumed to have followed a single ray path, and the multiple paths of the incident radiation to the receiver preclude simple assumptions about association of a transmission or reflection measurement with a single line integral. Also, a single frequency CW measurement is not capable of discriminating between tissue paths, the electrical length of which differ by more than one wavelength.
Other approaches which have been considered include pulsed radio frequency time delay and group delay measurement techniques. These approaches also suffer from several deficiencies. Pulse techniques require pulses with very fast rise time (100 picoseconds or less) in order to obtain adequate resolution. Such pulses require extremely broadband modulators, amplifiers, transmission lines, and detectors, which are complex and expensive. Group delay techniques are limited in the resolution which is obtainable by the noise and drift present in the network analyzers which must be utilized.
The present invention utilizes a time delay spectrum which is produced by linearly sweeping a microwave signal generator in frequency from a first frequency to a second frequency in a particular period. The generator output is divided into first and second signals which are propagated through a reference channel and a test channel, respectively, and the propagated signals are combined to produce a frequency spectrum which is representative of the differential time delay in the propagation of the signals through their respective channels.
A search of the prior art conducted in connection with the preparation of the present application developed the following patents in which a frequency varying signal and/or the combination of two signals are utilized in a measurement technique: U.S. Pat. No. 3,107,329 (McSkimin); U.S. Pat. No. 3,439,266 (Rogers); U.S. Pat. No. 3,851,244 (Mounce); U.S. Pat. No. Re. 29,008 (Ott). The McSkimin patent is an example of an application of pulse techniques and discloses a method wherein synchronism between two radio frequency pulses, one of which has been transmitted through a test channel, and one of which has been transitted through a reference channel, can be determined by comparison of the more centrally positioned portions of the pulses, so as to avoid the need for wide band apparatus and a high degree of resolution of each individual pulse. In accordance with one embodiment of the McSkimin method, the radio frequency of the pulses is continuously varied while the time interval between the pulses is varied until the signal produced by combining the pulses is at a substantially constant and clean maximum.
The Rogers patent discloses a heterodyne system for testing frequency sensitive electrical devices for insertion loss, return loss, and the like. The system is adapted to test devices which operate over a band of frequencies by the provision of a swept frequency generator which is coupled to a network having two transmission branches. One of the branches includes the device to be tested and the other branch includes an element which renders it electrically long with respect to the other branch. The outputs of the two transmission devices are heterodyned in a mixer to produce an intermediate frequency signal, which has an envelope indicative of a predetermined transmission characteristic of the device.
The Mounce patent discloses a microwave system of the heterodyne type for measuring moisture in sheet materials such as paper, wherein a detector is subjected both to strong signals from a frequency modulated source and to delayed and attenuated signals which have passed twice through the moisture containing material. The reason for doing so is that the strong signal acts as a local oscillator on the detector and forces it out of its high-impedance, high-noise, square-law region into a low-impedance, low-noise, linear region of operation.
The Ott patent discloses a machine for identification of persons wherein the frequency response characteristics of a portion of a person's body to sonic wave energy are compared with previously stored data for the person. Detector means are provided to detect the phase and amplitude of the signal transmitted through the person's body, and correlator means are provided to produce a transfer function representing an algebraic-trigonometric statement of the output divided by the input. In one embodiment a swept frequency source is employed and the output thereof may be applied to the correlator means for comparison with the signal transmitted through the person's body to obtain a transfer function having improved accuracy.
Applicants are also aware of U.S. Pat. No. 3,466,653 (Heyser), which discloses a time delay spectrometer for measuring the audio spectral response of an object which is located in an environment containing radiation reflecting bodies. In the Heyser device, a sound energy source is driven by a sweep frequency oscillator and the output of a sound detector is filtered by a tunable bandpass filter which is tuned or driven so as to follow the frequency of the sweep frequency oscillator by a delay equal to the time required for sound to travel a direct path from the source to the object and then to the detector. The filter thus passes only sound waves following the direct path and rejects sound waves which arrive at a later time when the filter has passed on to a new frequency. Heyser has also worked on ultrasonic imaging systems in which transmission variations through a target are measured. This work has been described in an article by Heyser and Le Croissette, 1 Ultrasound in Medicine and Biology 119-131.
Finally, applicants are aware of "CHIRP" radar techniques, wherein swept frequency radar signals are employed to reduce the peak power requirements of the radiated signal, and the delay time of the reflected signals is measured to determine target range. Reference is made to an article by Klauder, et al., appearing in Volume XXXIX of the Bell System Technical Journal (1960), at pages 745-808, for a discussion of CHIRP radar techniques.
The prior art cited hereinabove includes, in the opinion of the applicants, the closest prior art of which they are aware. However, there is no representation that no better art exists.
SUMMARY OF THE INVENTIONRemote interrogation of biological targets is accomplished in accordance with the present invention by time delay spectroscopy method and apparatus wherein a microwave signal is generated which varies in frequency from a first frequency to a second frequency in a predetermined time period. The signal is divided into two signals, one of which is propagated through a test channel comprising a transmitting antenna for transmitting the signal through the target, and a receiving antenna for receiving the signal transmitted through the target, and the other of which is propagated through a reference channel providing a fixed time delay of propagation. The propagated signals are detected and mixed to produce a time delay spectrum wherein the frequency of each spectral line represents the instantaneous difference in the frequencies of the detected signals. Preferably, the transmitting and receiving antennas and the target are immersed in water or other high dielectric medium.
In accordance with a further aspect of the invention the time delay spectrum is analyzed to determine the direct ray path through the target by selecting the spectral line having the highest amplitude at the lowest frequency.
Other features and advantages of the invention will be set forth in, or apparent from, the detailed description of a preferred embodiment found hereinbelow.
BRIEF DESCRIPTION OF THE DRAWINGSFIG. 1 is a simplified block diagram of a time delay spectrometry system constructed in accordance with the present invention for remote interrogation of a biological target.
FIG. 2a is a front elevational view, partially cut away, of a perferred antenna embodiment for use in the system of FIG. 1.
FIG. 2b is a side elevational view, of the antenna shown in FIG. 2a.
FIG. 3 is a graphical representation of an illustrative time delay spectrum produced by a system similar to that shown in FIG. 1 with an antenna of the type shown in FIGS. 2a-2b.
FIG. 4 is a graphical representation of the time delay spectrums produced by a system similar to that shown in FIG. 1 during calibration thereof.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTSReferring to FIG. 1, asystem 10 constructed according to the present invention for remote interrogation of a biological target comprises a sweptmicrowave oscillator 20 which is linearly swept by asweep generator 22 to produce a microwave signal which continuously varies in frequency from a first frequency f1 to a second frequency f2 in a predetermined time period TS.
System 10 further comprises apower divider 30 for dividing the output ofoscillator 20 into first and second propagation signals which are propagated, respectively, through atest channel 32 and areference channel 34.Test channel 32 comprises a transmittingantenna 36, a biological target A situated in free space, generally denoted 37, and a receivingantenna 38.Reference channel 34 provides a signal path having a fixed time delay.
Preferably, the biological target A, as well asantennas 36 and 38, are immersed in water, or other high dielectric constant environment. There are several advantages to a water environment with respect to remote interrogation of a biological target. First, the wavelength of the propagated signal is contracted, which improves the spatial resolution of the line scan. Second, coupling of energy into the target is improved. Third, discrimination of multipaths through the target is facilitated. Fourth, water has acceptable loss, inertness, and tissue match characteristics for biological targets.
System 10 also comprises aproduct detector 40 for detecting and multiplying, or mixing, the first and second propagation signals following propagation thereof through the correspondingchannels 32 and 34 to produce signals representing the time delay spectrum. In the case of sinusoids, the result of the mixing is sinusoids whose frequencies are the sum and difference, respectively, of the frequencies of the input signals. Alow pass filter 50 andFourier analyzer 60 are provided for processing of the time delay spectrum.Filter 50 rejects the spectrum signal whose frequency is the sum of the frequencies of the inputs todetector 40, and also acts as an anti-aliasing filter to reduce errors in the digital Fourier transform produced byanalyzer 60.Analyzer 60 digitizes the output offilter 50 and transforms the data to amplitude and phase information in the frequency domain.
Antennas 36 and 38 should preferably have as large a bandwidth, and as small an aperture, as possible, with a phase center that does not move with frequency. The need for a large bandwidth is a consequence of the swept frequency nature of the interrogating radiation. For a particular measurement, the bandwidth required is determined by the resolution needed, the sweep rates available inoscillator 20, the sweep range available, and the sampling rates provided by theanalyzer 60 which is utilized. A small aperture is desirable since if the aperture is too large compared to the area of the dielectric discontinuity being studied, it is not possible for the discontinuity to occlude the direct ray path betweenantennas 36 and 38, and a condition equivalent to multipath occurs which obscures the desired result. A constant phase center is preferred since movement of the phase center with frequency is equivalent to varying the separation between antennas, which has the effect of broadening the spectral lines in the time delay spectrum. A preferred antenna construction is depicted in FIGS. 2a-2b. The antenna, generally denoted 80, comprises a doubleridged waveguide 82, which is approximately 6.7 mm in length. The length ofwaveguide 82 represents a compromise between internal loss and ease of impedance matching, since a shorter length, on the order of 3 mm, would be preferable from the standpoint of power loss, but would not permit the use of tuning screws for impedance matching.
The top wall ofwaveguide 82 is provided with anaperture 83, through which thefeed probe 84 of a standard 50 ohm impedancecoaxial input cable 86 having a fluorocarbon dielectric such as "Teflon" is inserted by means of a standardfemale connector 85 mounted on the top surface ofwaveguide 82 and a standardmale connector 87 mounted on the end ofcable 86.
As shown, thetop ridge 90 ofwaveguide 82 extends longitudinally along the upper interior surface only from the front end ofwaveguide 82 to the perimeter ofaperture 83, while thebottom ridge 91 extends longitudinally along the entire lower interior surface from the front end to the rear end, ofwaveguide 82. In addition, as shown, the front ends of bothtop ridge 90 andbottom ridge 91 are bevelled. Twoholes 94 and 95, size 2-56, are provided in bottom ridge of 91 ofwaveguide 82 for receiving tuning screws 96 and 97, respectively, which are used to obtain a broader impedance match.Hole 94 is substantially coaxial withaperture 83. It is also noted that therear screw 96 does not protrude intowaveguide cavity 88, whilefront screw 97 does protrude intocavity 88.
Feed probe 84 is inserted intocavity 88 and is shorted to tuningscrew 96, and thus to thebottom ridge 91 ofwaveguide 82 in order to control the VSWR ofantenna 80. Preferably,probe 84 is oriented substantially perpendicularly with respect tobottom ridge 91. The diameter ofprobe 84 is reduced to approximately 0.5 mm to provide a better match to the high impedance ridges ofwaveguide 82.Antenna 80 further comprises a shortingplate 92 mounted at the rear ofwaveguide 82 and positioned with respect to feedprobe 84 so as to obtain the smoothest impedance match over the operating bandwidth of the antenna. Shortingplate 92 is provided with a 2.2mm diameter hole 93 to facilitate removal of air bubbles trapped inwaveguide 82 whenantenna 80 is immersed in the dielectric medium, and to permit alignment ofantenna 80 with respect to the target. The dimension ofhole 93 is determined by the bandwidth of the radiation to be transmitted, being sized so as to be below the cutoff frequency for the bandwidth of the radiation.
The dielectric offeed probe 84 is preferably inserted intoaperture 83 such that the dielectric is approximately even with the upper interior surface ofwaveguide 82. Final impedance matching is obtained by simultaneous adjustment of tuningscrews 96 and 97 and penetration of the dielectric intoaperture 83.
Antenna 80 is advantageously enclosed in a conventional doubleridged waveguide flange 99, which provides mechanical stability and means for mounting extensions ontoantenna 80. Preferably,flange 99 is machined withnotches 100 and 101 to permit connection offeed cable 86 ontoconnector 85, and access to tuningscrews 96 and 97.
In use,antenna 80 and the associatedflange 99 are advantageously mounted at the end of ahollow tube 102 which supportsantenna 80 and provides a conduit which protectscable 86.Cable 86 is terminated at the distal end oftube 102 in a conventionaltype N connector 104. To reduce the effect of reflection offtube 102,antenna 80 is supported 5 cm in front oftube 102 by means of a metal standoff 106 andconnector 107.Antenna 80 is designed to be operated totally immersed in the dielectric medium and have an operating bandwidth of 2000 MHF to 4000 MHF. The dimensions which have been cited hereinabove assume that the dielectric medium is water, which is preferably distilled and at a temperature of 32 C. If a medium with a different dielectric constant is to be used, then the dimensions would need to be altered accordingly. In general, if the medium has a dielectric constant lower than that of water, larger dimensions would be required, and conversely, if the medium has a dielectric constant higher than that of water, smaller dimensions would be required.
The various other components ofsystem 10 can be conventional. For example, a Hewlett-Packard 8690B oscillator, Hewlett-Packard 8699 sweep generator, Microlab/FXR Model DA-2FN power divider, Watkins-Johnson Model M1G product detector, Krohn-Hite Model 3343 filter, and a Hewlett-Packard 5451B analyzer may be employed foroscillator 20,generator 22,power divider 30,product detector 40,filter 50, andanalyzer 60, respectively.
It is also to be noted thatsystem 10 may also advantageously include various well-known and conventional devices, such as attenuators, filters and the like which are commonly included in any microwave system to improve the operating characteristic thereof. Since such devices are not an aspect of the present invention, a detailed description thereof has been omitted in the interests of clarity.
Turning to the operation ofsystem 10, when the time delay throughtest channel 32 andreference channel 34 is not the same, the instantaneous frequencies of the signals detected bydetector 40 will be different, and the magnitude of the frequency difference will be inversely proportional to the sweep time Ts, and directly proportional to the difference in propagation times in thechannels 32 and 34. Further, the amplitude of the direct path signal inchannel 32 detected bydetector 40 is a function of the attenuation through the target. More specifically, the general expression for the signal produced at the output offilter 50 resulting from a single ray path is as follows:
E = (αβA.sup.2 /2) cos[k.sub.f Tt - ω.sub.1 T - k.sub.f T.sup.2 ] [U(t-T) - U(t-T.sub.s)]
Where
ω1 = 2πf1 = Start frequency (radians/second)
ω2 = 2πf2 = Stop frequency (radians/second)
Ts = Sweep time (seconds)
T = Differential time delay between test and reference channels (seconds)
t = Time (seconds)
α = Attenuation through the path under study
β = Multiplier which is a function of the gain of the measurement system
A = Amplitude of the signal at the power divider output
kf = (ω2 - ω1)/Ts Radians/second2
The function U(x) is defined as follows:
U(x)=0 for x<0
U(x)=1 for x≧0
The result, for a single path, is a truncated sinusoidal waveform with an angular frequency equal to kf T radians/second and with a phase offset equal to (ω1 T + kf T2) radians.
If appropriate values for the above identified are chosen, such as for example:
Ts = 7.75 milliseconds; f1 = 2.0 GHz; and f2 = 4.0 GHz; the ω1 T term dominates the phase offset until the differential time delay T becomes very large (equivalent to a free space propagation distance of over 100 miles). Thus, for biologically relevant values of differential time delay, the time delay spectrum for a single path may be written as
E' = (αβA.sup.2 /2) cos [k.sub.f Tt - ω.sub.1 T] [U(t-T) - U(t-T.sub.s)]
For delays T less than the sweep time Ts, examination of the spectrum of the signal at the output ofproduct detector 40 yields an unambiguous measure of time delay through the target. Furthermore, since time of propagation through a particular path in the target uniquely determines the frequency of a spectral line in the time delay spectrum, multiple paths can be resolved on the basis of time-of-arrival information.
FIG. 3 illustrates a typical time delay spectrum produced at the output offilter 50 in asystem 10 utilizing smallapertured antennas 36 and 38 of the type illustrated in FIGS. 2a-2b. Thewaveform 70 constitutes the envelope of the individual spectral lines, which each represent the instantaneous difference in frequencies of the signals detected bydetector 40, and thus represents the time delay for a particular path of propagation intest channel 32. Thewaveform 70 thus represents a composite time delay spectrum for all paths of propagation intest channel 32. The direct ray path through the target corresponds to the spectral line having the highest amplitude at the lowest frequency, e.g. point 70A onwaveform 70 in FIG. 3. Withantennas 36 and 38 of sufficiently small aperture, such as antennas of the type illustrated in FIGS. 2a-2b, the resolution of the direct ray path is readily accomplished since the time delay spectrum peaks sharply, as shown in FIG. 3, at the spectral line corresponding to the direct ray path. If the apertures ofantennas 36 and 38 are not sufficiently small, resolution of the direct ray path is in principle still possible. However, additional analysis of the complex time delay spectrum which results is required, which becomes too involved to have practical utility, at least with respect to interrogation of biological targets, as the size of the antenna aperture increases. Once thesystem 10 is calibrated, as described in more detail hereinbelow, the time delay through the target may be determined directly from the frequency of the direct ray path spectral line. It is to be noted that the value of the time delay of the signal path provided byreference channel 34 may be chosen such that for a given biological target and environment, the output ofproduct detector 40, and hencewaveform 70, are in a frequency range, for example, the audio range, which is easily processed.
System 10 is calibrated by determining the propagation delays caused by the elements oftest channel 32 other than the target A. These delays may be either calculated theroretically, or determined empirically. Referring to FIG. 4, one method of empirically calibratingsystem 10 involves producing atime delay spectrum 72 withsystem 10 in its normal configuration. An element having a known time delay, e.g. a predetermined length of RG9/U coaxial cable, is inserted intotest channel 32 betweenpower divider 30 andantenna 36, and anothertime delay spectrum 74 is produced. The difference between the expected shift in the location ofspectrum 74 with respect tospectrum 72, based on the time delay of the introduced element, and the actual shift which occurs provides a measure of the time delay introduced by the elements oftest channel 32.
Tests conducted in connection with the preferred embodiment of the invention described hereinabove demonstrate that the measured data does not become ambiguous until the propagation time is equivalent to that of over 1400 miles in free space, which is manifestly adequate for biologically relevant targets. Additionally, the system is capable of discriminating between two ray paths whose differential time delay is on the order of 100 picoseconds, which corresponds to a differential path length of 6mm in brain tissue and the like.
Although the invention has been described with respect to an exemplary embodiment thereof, it will be understood that variations and modifications can be effected in the embodiment without departing from the scope or spirit of the invention.